In this particular case, OP said none of the others met their needs. I would like to know what new functionality this one has to know if it’s something I’m interested in or not. It’s not a critique - it’s helping me understand if I want to check it out or not.
His model actually accounts for whether polls were taken before or after an event, and raises and lowers their impact and error margin based on that. Right up to the debate, his model was giving Kamala a <30% chance and it’s only the inclusion of new polls since the debate that have moved her to 50%.